Search

IPIN 2019 Competition, sixth in a series of IPIN competitions, was held at the CNR Research Area of Pisa (IT), integrated into the program of the IPIN 2019 Conference. It included two on-site real-time Tracks and three off-site Tracks. The four Tracks presented in this paper were set in the same environment, made of two buildings close together for a total usable area of 1000 m2 outdoors and and 6000 m2 indoors over three floors, with a total path length exceeding 500 m. IPIN competitions, based on the EvAAL framework, have aimed at comparing the accuracy performance of personal positioning systems in fair and realistic conditions: past editions of the competition were carried in big conference settings, university campuses and a shopping mall. Positioning accuracy is computed while the person carrying the system under test walks at normal walking speed, uses lifts and goes up and down stairs or briefly stops at given points. Results presented here are a showcase of state-of-the-art systems tested side by side in real-world settings as part of the on-site real-time competition Tracks. Results for off-site Tracks allow a detailed and reproducible comparison of the most recent positioning and tracking algorithms in the same environment as the on-site Tracks., The authors would like to thank Siemens, the Electronic and Telecommunications Research Institute (ETRI), the Korean Institute of Communications and Information Sciences (KICS), Ganko Food, J-Power Systems, and the other sponsors who provided prizes for the winners of the competition (see http://ipin-conference.org/2019/awards.html for a complete list).

Vehicular Ad hoc Networks (VANETs) are the underlying network for all the communications between Intelligent Transportation System entities as, for example, Road Side Units, On-Board Units, etc. These networks are growing in size and interest from the automaker's industry, software companies and scientific researchers. The intrinsic technological nature of VANETs (based on a wireless shared medium) with no predefined structure of architecture (high mobility of their users) makes them very susceptible to security attacks, which may impact deeply on road users. There are several works on VANET security, usually involving cryptography. But, these can only prevent some types of attacks. Others, mainly the ones carried by authentic entities, cannot be prevented applying this strategy. Nonetheless, attacks can be detected triggering responses minimizing the repercussions. This can be accomplished by using an Intrusion Detection System (IDS) which can be enhanced with the help of Machine Learning (ML). But, one of the main difficulties for an effective implementation of a ML-based IDSs is the (un)availability of valid public datasets. The IDSs need a large collection of VANET messages to be able to recognize types of behaviors. However, most of the studies in the literature discussing solutions and their results do not make the datasets publicly available for scrutiny by third parties. As such, the results are difficult to properly verify and validate. The goal of this work is to provide well documented and publicly available datasets to be used by researchers on future RD projects, enabling a fairer comparison of their results., This work has been supported by national funds through FCT - Fundacao para a Ciéncia e Tecnologia within the Project Scope: UIDB/00319/2020 and by the European Structural and Investment Funds in the FEDER component, through the Operational Competitiveness and Internationalization Programme (COMPETE 2020) [Project no 039334; Funding Reference: POCI-01-0247-FEDER-039334].

Previous studies have shown that the ITS solution called platooning allows the autonomous collaborative driving and can improve traffic safety and throughput. Traffic flow is optimized by Cooperative Adaptive Cruise Control (CACC), which allows for the automatic short-distance vehicle following, using inter-vehicle wireless communication in addition to onboard sensors. This paper presents the platooning vehicle longitudinal controller evaluation using simulation environment. The employed controller uses IEEE 802.11p technology for vehicle-to-vehicle (V2V) communications on Vehicular Ad hoc Network (VANET). To evaluate the CACC the Veins simulation framework was used and the complete simulation setup is described in this paper. The presented analysis expound the methodology to verify controller safety and stability characteristics within the different traffic scenarios and platooning maneuvers using the simulation., This work was sponsored by the Portugal Incentive System for Research and Technological Development. Project in copromotion no. 002797/2015 (INNOVCAR 2015-2018), and also by COMPETE: POCI-01-0145-FEDER-007043 and FCT - Fundação para a Ciência e Tecnologia within the Project Scope: UID/CEC/00319/2013.

A Vehicular Adhoc Network (VANET) is a conceptualisation that can be applied to the communications domain of an Intelligent Transportation System (ITS), where vehicles use On Board Units (OBU) to communicate with other vehicles or the road side infrastructure. The heterogeneity of the interconnected devices, the large number of devices inside and outside the vehicles and different degrees of device density, makes interoperability a challenging problem, making way to many machine-2-machine (M2M) autonomous communications. This paper discusses two different M2M architectures that may be used in ITS context and presents test results obtained from implementations of these M2M architectures on real OBU hardware, in the context of ETSI compliant safety applications.The experimental results were taken using two different oneM2M Vehicular Architectures Types in order to analyse their suitability to be used in a platooning car-following use case. For the car-following use case, OBUs interface the vehicle's Controller Area Network (CAN) bus and use Dedicated Short Range Communications (DSRC) for M2M communications to convey dynamic data to enable controlling the vehicle's speed and longitudinal distance. The experiments and corresponding result analysis has shown that the oneM2M architectures are suitable to be used for such car-following application, which requires latencies below 100ms., This work has been sponsored by the Portugal Incentive System for Research and Technological Development. Project in co-promotion 002797/2015 (INNOVCAR 2015-2018), and also by COMPETE: POCI-01-0145-FEDER-007043 and FCT - Fundação para a Ciência e Tecnologia within the Project Scope: UID/CEC/00319/2019.

Platooning is an emerging Intelligent Transportation Systems (ITS) application that can have a huge impact on road traffic. It allows vehicles to travel in a convoy manner, very close to each other with constant speeds and gaps.It uses Vehicular Ad hoc Networks (VANETs) as the underlying communication framework. These are a type of network with a set of very specific characteristics, making it an easy target for attackers. Frequent disconnections, topology change, communication over the air among others, create vulnerabilities easily exploitable.The goal of this paper is to present how to secure the communication between the several nodes of a platooningby using an existing security model. After a careful research and evaluation, the Vehicular Ad hoc Network Public Key Infrastructure and Attribute Based Encryption with Identity Manager Hybrid (VPKIbrID) model was chosen. It provides a whole solution with tools for secure message exchange and entity authentication.The impact of the introduced security model was evaluated using simulations with different platoon sizes and VPKIbrID modes. The several operations of the platoon will be carried using the Platoon Management Protocol (PMP).The result analysis allowed to verify that when using a VPKIbrID mode that allows multicast/broadcast communications, the vehicles sent much fewer messages than the mode that only supports unicast. The performance results indicate that the vehicles should use VPKIbrid Public Key Infrastructure (VPKIbrID-PKI) for maneuvers and VPKIbrid Attribute Base Encryption (VPKIbrID-ABE) to exchange platoon beacons or any message exchanged to all platoon members. Moreover, the messages broadcasted in a platoon group will always have the same targets, with the same attributes, being a perfect use case for the usages of cached keys., This work has been sponsored by the Portugal Incentive System for Research and Technological Development. Project in co-promotion no 002797/2015 (INNOVCAR 2015-2018), and also by COMPETE: POCI-01-0145-FEDER-007043 and FCT - Fundacao para a Ciencia e Tecnologia within the Project Scope: UID/CEC/00319/2013.

Platooning is a solution that allows automated vehicles to travel very close to each other, enabling enhancements in terms of safety, traffic flow and highway capacities. Furthermore, it provides a more convenient and comfortable driving experience to the drivers. The research on Platooning covers a very large spectrum of areas, ranging from communication strategies to controller solutions. However, most of the proposed works tend to present very specific solutions, treating platooning almost as a "black-box". Thus, it is imperative to find solutions that allow a proper assembly of every component that is crucial for a proper behavior of the system as a whole. This paper presents an open and modular architecture for platooning as a service, discussing its general assumptions and the logical and functional models. Additionally, some related preliminary results are also presented., This work has been sponsored by the Portugal Incentive System for Research and Technological Development. Project in co-promotion no 002797/2015 (INNOVCAR 2015-2018), and also by COMPETE: POCI-01-0145-FEDER-007043 and FCT - Fundacao para a Ciencia e Tecnologia within the Project Scope: UID/CEC/00319/2013.

Nowadays it is common for vehicles to be equipped, not only with huge computing power, but also with great communication capabilities, materialized in multiple and various wired and wireless interfaces. This paper surveys the most common inter and intra vehicle technologies. First, an introduction to Intelligent Transportation Systems and Vehicle Communication Systems is presented. Then, the technologies usually found in vehicles are characterized in detail - their advantages and flaws regarding particular communication scenarios are also discussed. Finally, the paper concludes with a mapping between potential wireless communication technologies for the different types of ITS applications., This work has been supported by COMPETE: POCI-01-0145-FEDER-007043 and FCT - Fundação para a Ciência e Tecnologia ˆ within the Project Scope: UID/CEC/00319/2013

Studies in Computational Intelligence, 616, Recent advanced intelligent devices enable vehicles to retrieve information while they are traveling along a road. The store-carry-and-forward paradigm has a better performance than traditional communication due to the tolerance to intermittent connectivity in vehicular networks. Named Data Networking is an alternative to IP-based networks for data retrieval. On account of most vehicular applications taking interest in geographic location related information, this paper propose a Geographical Opportunistic Forwarding Protocol (GOFP) to support geo-tagged name based information retrieval in Vehicle Named Data Networking (V-NDN). The proposed protocol adopts the opportunistic forwarding strategy, and the position of interest and trajectories of vehicles are used in forwarding decision. Then the ONE simulator is extended to support GOFP and simulation results show that GOFP has a better performance when compared to other similar protocols in V-NDN., This work is supported in part by the Fundamental Research Funds of Jilin University, No. 450060491509 and partially supported by FCT-Fundacao para a Ciencia e Tecnologia Portugal in the scope of the project: UID/CEC/00319/2013.

Devido às características próprias das redes móveis ad hoc (Mobile Ad hoc Network - MANET), dotar este tipo de redes de garantias de qualidade de serviço (QoS) no tráfego fim a fim torna-se um desafio. Este artigo apresenta um protocolo de encaminhamento com QoS para redes ad hoc, que se designa por Ad hoc QoS Multipath Routing with Route Stability (QMRS), que tem como objectivo suportar aplicações com requisitos de qualidade de serviço, nomeadamente requisitos no atraso fim a fim. Este protocolo tem a possibilidade de encontrar até três rotas de nós disjuntos que cumpram o requisito de QoS. Adicionalmente e com o objectivo de garantir a estabilidade do processo de encaminhamento, usa a potência de sinal das ligações entre nós vizinhos para eleger a rota mais estável, rota essa que passa a ser usada para o reenvio do tráfego. Quando se verifica a existência de rotas com uma estabilidade idêntica, dá-se preferência à rota com menor atraso fim a fim. O protocolo detém também um mecanismo de manutenção, recuperação e verificação de incumprimento do requisito de QoS nos caminhos encontrados. Os resultados obtidos na simulação realizada permitem verificar, que o protocolo QMRS implementado, reduz o atraso fim a fim e aumenta a taxa de entrega de pacotes no destino, comparativamente com protocolo Adhoc On-Demand Distance Vector (AODV)., Fundação para a Ciência e a Tecnologia (FCT)

A RCTS tem em operação, desde há vários anos, um conjunto de sondas do tipo appliance destinadas a aferir de forma contínua a qualidade da rede, tanto em IPv4 como IPv6. Normalmente, pelo facto de serem appliances, consequentemente “fechadas”, cria-se uma dependêndia de hardware e software. Apesar das appliances possuírem algumas características únicas ao nível de hardware, podem desenvolver-se soluções abertas (open source) equiparadas. Na realidade, vários sistemas de aferição de desempenho de redes, desenvolvidos em open source, tanto na GÉANT (rede académica europeia) como na Internet2 (rede académica americana), atingiram uma notoriedade que permite considerá-los como alternativas credíveis para as atuais sondas, se se provar a sua maturidade e estabilidade. Neste trabalho apresenta-se o desenvolvimento e teste de uma especificação de uma nova sonda para a RCTS, baseada em soluções open source., FEDER, através do Programa Operacional Fatores de Competitividade - COMPETE - e por Fundos Nacionais através da FCT - Fundação para a Ciência e Tecnologia, através do Projecto: FCOMP-01-0124-FEDER-022674

Delay Tolerant Networks (DTNs) aim to increase messages delivery ratio in environments where it is not possible to establish an end-to-end connection. Although the research of new DTN routing protocols has been gaining some relevance, those protocols usually assume that nodes in a network will collaborate. Nodes can behave selfishly, leading to the inappropriate use of resources, following up the malfunction of the network environment. This paper presents an extension based on bayesian game theory to existing routing protocols. Each node tries to figure others node’s type using the Naive Bayes classifier and behaves appropriately in order to achieve optimal results across the cooperative nodes. The regarded data through the exchangeable events between nodes can also be used to calculate each node’s selfishness, assigning the acceptance and respective delivery probability of a message to its destination. The filter extension improved the delivery ratio of the cooperative nodes on selfish networks., FEDER Funds through the Programa Operacional Fatores de Competitividade COMPETE and by National Funds through the FCT - Fundação para a Ciência e a Tecnologia (Portuguese Foundation for Science and Technology) within project FCOMP-01-0124-FEDER-022674

Large-scale pervasive public displays networks are becoming an emerging paradigm and represent a radical transformation in the way we think about information dissemination in public spaces. One of the features of pervasive public display systems is their ability to create experiences that span across multiple displays in a coordinated fashion. Proprietary single site display solutions exist but these are not open to third-party developers. On the other hand, scalable open systems that enable large-scale, synchronised and multi-screen experiences, spanning multiple networks domains will call for the definition of multiple administrative boundaries that accommodate function partitioning. In our research, we are studying the key requirements involved in this open application synchronisation and present our initial work on designing a synchronisation model and Application Programming Interface for public displays application developers that is built on top of the PubSubHubbub protocol, an open protocol for distributed publish/subscribe communication on the Internet., Fundação para a Ciência e a Tecnologia (FCT)

Large-scale pervasive public displays networks are becoming an emerging paradigm and represent a radical transformation in the way we think about information dissemination in public spaces. These networks with its pervasive nature rise a number of challenges for those who have to design, test, deploy and use them. It is imperative to understand what are the key tradeoffs in the design of pervasive displays networks, mainly on their components and respective protocols, in order to provide a fully open, global and most importantly scalable displays network.

O conceito de virtualização de redes encontra-se em expans. A capacidade de oferecer a uma infra-estrutura de rede a possibilidade de suportar várias redes virtuais distintas, sem as pressões impostas pelas redes de suporte,é um ponto de interesse tanto para operadoras como para académicos. Esta característica permite criar um ”mundo” virtual, auto-contido e com parâmetros controlados e alteráveis onde podem ser realizadas experiências com novas tecnologias ou mesmo, olhando sob um ponto de vista empresarial, oferecer serviços específicos a clientes, onde os parâmetros da rede são contratualizados com um revendedor. Este artigo descreve uma proposta de implementação de rede virtual, onde se aborda a descoberta de recursos e as diferentes metodologias para a troca de informacão entre os nós virtuais. As ligações criadas permitem virtualizar uma arquitectura níıvel 2 considerando cenários de conexão entre dois ou mais nós. São utilizadas várias abordagens de modo a criar a rede virtual, nomeadamente conexões baseadas em Sockets e RawSockets, assegurando uma comunicação controlada sobre uma infra-estrutura de rede operacional. Os parâmetros que estas ligações permitem oferecer foram avaliados numa situação real comprovando a possibilidade de disponibilização desta tecnologia sobre a infra-estrutura pública.

As redes Peer-to-peer (P2P) tem vindo a ganhar popularidade na internet, devido ao aumento do número de utilizadores e de serviços de natureza distribuída, como a partilha de ficheiros, e as chamadas de voz sobre IP (VoIP). A criação deste tipo de redes, baseando-se em soluções abertas como o SIP, pode facilitar a criação de novos tipos de serviços, assim como permitir uma mais fácil integração de diferentes serviços. Neste trabalho foi desenvolvida uma implementação JAVA, capaz de criar redes P2PSIP com um ou dois níveis hierárquicos. A comunicação entre os nós da rede P2P, ´é feita através de um protocolo totalmente baseado em SIP. Como algoritmos a utilizar pelo overlay P2P, foram implementados o algoritmo Chord e EpiChord. Para comprovar o funcionamento da implementação JAVA, foram efectuados testes num ambiente real, recorrendo a uma topologia de rede emulada com o CORE.

In last years, many efforts have been made to build indoor positioning systems using different types of signals and different types of position algorithms. Experiences with sound, light, infra-red, Bluetooth and popular IEEE 802.11 network signals have been carried out both by commercial and academic entities. The major goal is to fulfill the lack of GPS coverage inside buildings with a more precise and/or cheaper alternative. WiFi based Localization systems gain special attention in this context due to its huge usage inside all modern facilities. Many algorithms have been proposed to estimate the device position within the WiFi network using the signal strengths (RSS) obtained by 802.11 device drivers. RSS values can be registered by each packet received, either by the the user’s device or by the network’s Access Points. Values can then be compared with a pre-configured map of signal strengths obtained in a calibration (offline) phase. This paper presents a localization system, where the location is determined exclusively by the network infrastructure. The location computation is based on the signal strength measured at Access Points and sent to a localization server. The localization server determines the location, stores it in a database and accepts requests from authenticated clients. Results obtained in experimental tests are discussed and compared with those published by other authors, allowing us to conclude that the system combines a good accuracy, with the advantage of not requiring any intervention from clients in the localization process. The entire system is designed in order to make locations available as a network service.

Para dotar o encaminhamento Inter-domínio, baseado no BGP, da capacidade de determinar caminhos dependendo dos requisitos de Qualidade de Serviço do tráfego fim a fim, torna-se fundamental incorporar métricas de QoS nas mensagens BGP trocadas entre os Routers Fronteira dos Sistemas Autónomos (AS). Esta proposta apresenta a implementação de uma estratégia de encaminhamento Inter-domínio com QoS (riQoS), desenvolvida sobre uma extensão BGP ao NS2 (ns-BGP). O riQoS estende os atributos do BGP e utiliza as mensagens UPDATE para divulgar informações de QoS entre os vários Routers Fronteira dos ASs. Para isso, o riQoS implementa alterações nos processos de comparação de rotas e de decisão do BGP tendo em conta a Largura de Banda e o Atraso em cada um dos Routers e ligações entre si. Em vez de utilizar métricas determinísticas, o BGP anuncia as informações através de métricas com significado estatístico: Available Bandwidth Index (ABI) e Delay Index (DelayI) [1], que permitem estimar os valores instantâneos da Largura de Banda disponível e do Atraso fim a fim.

This paper addresses the problem of designing a suitable Quality of Service (QoS) solution for mobile environments. The proposed solution deploys a dynamic QoS provisioning scheme able to deal with service protection during node mobility within a local domain, presenting extensions to deal with global mobility. The dynamic QoS provisioning encompasses a QoS architecture that uses explicit and implicit setup mechanisms to request resources from the network for the purpose of supporting control plane functions and optimizing resource allocation. Abstract--- For efficient resource allocation, the resource and mobility management schemes have been coupled resulting in a QoS/Mobility aware network architecture able to react proactively to mobility events. Both management schemes have been optimized to work together, in order to support seamless handovers for mobile users running real-time applications. Abstract--- The analysis of performance improvement and the model parametrization of the proposed solution have been evaluated using simulation. Simulation results show that the solution avoids network congestion and also the starvation of less priority DiffServ classes. Moreover, the results also show that bandwidth utilization for priority classes is levered and that the QoS offered to Mobile Node's (MN's) applications, within each DiffServ class, is maintained in spite of MN mobility. Abstract--- The proposed model is simple, easy to implement and takes into account the mobile Internet requirements. Simulation results show that this new methodology is effective and able to provide QoS services adapted to application requests.

Wireless networks will certainly run applications with strict QoS requirements and so, micro-mobility protocols such as fast hierarchical mobile IPv6 (FHMIP) are useful tools to accomplish this new feature. The FHMIP is an effective scheme to reduce Mobile IPv6 handover disruption, however it does not support application's QoS requirements. Therefore, in order to provide QoS guarantees for real-time applications it is necessary to develop new traffic management schemes; this implies the optimization of network mobility support and also some network congestion control. A traffic management scheme of this type should take into account the QoS requirements of handover users and should implement a resource management (RM) scheme in order to achieve this. In this paper, a new RM scheme for the DiffServ QoS model is proposed. This new scheme is implemented by access routers as an extension to FHMIP micromobility protocol. In order to prevent QoS degradation of the existing traffic, access routers should evaluate the impact of admitting a new mobile node (MN), previously to the handover. This evaluation and sequent decision on wether admitting or refusing MN's traffic is based on a measurement-based admission control (MBAC) algorithm. This architecture, that has been implemented and tested using ns-2, includes a simple signaling protocol, a traffic descriptor and exhibits an adaptive behavior to traffic QoS requirements. All the necessary measurements are aggregated by class-of-service, thus avoiding maintaining state on the individual flows., (undefined)

Tese de doutoramento em Informática, A maioria das propostas para encaminhamento multicast com QoS propõem que as rotas sejam calculadas a pedido, sendo necessário que os pedidos explicitem os recursos de que necessitam. Como resultado, neste contexto, o objectivo do encaminhamento com QoS é satisfazer pedidos individuais de requisitos, deixando para os mecanismos de reserva de recursos a responsabilidade de manter os recursos necessários à satisfação desses requisitos depois de ter sido descoberto, pelo processo de encaminhamento com QoS, um caminho exequível. Este tipo de estratégia adequa-se ao modelo DiffServ, mas não parece adequado na presença de redes DiffServ. De acordo com o modelo DiffServ, os diferentes fluxos de tráfego são agregados em classes de serviço e cada fluxo recebe um tratamento específico de acordo com a classe de serviço a que pertence. Deixam de ser dadas garantias por fluxo passando existir apenas diferenciação por classe de serviço. Nestas redes DiffServ devem então ser calculadas rotas por classe de serviço e não por fluxo como até aqui. Esta tese propõe um protocolo de encaminhamento multicast, o MCMRP (Multi Class Multicast Routing Protocol), um novo protocolo que permite a construção de árvores de distribuição de tráfego por classes de serviço. As heurísticas propostas permitem o esta- belecimento de árvores directas, em vez das árvores invertidas usadas habitualmente pelos protocolos de encaminhamento rnulticast tradicionais, relevando a importância das assimetrias das ligações num ambiente que é essencialmente unidireccional. Um dos principais elementos deste protocolo de encaminhamento multicast é o protocolo de encaminhamento unicast subjacente, que terá obrigatoriamente que ser capaz de encontrar uma rota por cada classe de serviço considerada, uma vez que para construir um novo ramo por cada classe de serviço, o MCMRP consulta a tabela de encaminhamento unicast. Para satisfazer esta nova necessidade, foi também concebido e implementado um novo protocolo de encaminhamento unicast por classes de serviço, o CoSLSP (Class of Service Link State Protocol). Ambos os protocolos foram implementados usando o Network Simulator. Os resultados obtidos através das simulações foram analisados e comparados com os de outros protocolos. A principal hipótese subjacente a este trabalho é a de que, nas redes DiffServ, uma estratégia de encaminhamento por classes de serviço pode complementar eficientemente, ao nível do encaminhamento, a diferenciação de tráfego efectuada ao nível dos nós., Most of current multicast QoS routing proposals are based on the principle that QoS routes must be computed for each request, where requests explicitly express their resource requirements. As a result, with this environment, the goal of QoS routing is to satisfy individual request requirements, resorting to resource reservation mechanisms to maintain those requirements after a feasible path has been found. This type of strategy is suited within the IntServ model but does not seem adequate in presence of DiffServ networks. According to DiffServ model, traffic flows are aggregated into specific classes-of-service and each flow receives a specific treatment accordingly to its class-of-service. There are no per flow guarantees, only per class differentiation. In this environment instead of per flow path computation, per class path calculation should be made, and so, within multicast scenarios, multiple multicast trees must be computed in order to satisfy different QoS requirements of different traffic classes. This thesis proposes MCMRP (Multi Class Multicast Routing Protocol), a new multicast routing protocol enabling per class multicast tree computation. The proposed heuristics enable directed distribution trees establishment, instead of reverse path ones, due to the importance of link asymmetry within an environment which is, essentially, unidirectional. One of the major elements of this multicast routing protocol is the unicast routing protocol in use. Although MCMRP is independent of the underlying unicast routing protocol, it must be a multi-class enabled unicast routing protocol. In other words, the unicast routing protocol must be able to find the unicast routes that can meet the QoS requirements of each Class of Service. III order to build a new tree branch for each Class of Service the multicast routing protocol will search the unicast routing table for the unicast path that is more adequate to satisfy the Q0S requirements of each class. To accomplish this new feature, a new unicast routing protocol called CoSLSP (Class of Service Link State Protocol) is proposed and implemented. Both protocols are implemented and simulated using Network Simulator. A set of simulation results are presented, analyzed and compared. The main assumption behind this work is that a per class path computation can complement, at routing level, the node level differentiation techniques, in the provision of per class differentiated handling.

Future Internet applications are expected to have higher QoS requirements. Therefore, routing protocols must be adapted to find a path which satisfies such requirements. Most multicast applications are QoS sensitive in nature, because they involve the transmission of real-time multimedia data streams. There are several research projects aiming to extend existing multicast routing protocols with QoS capabilities, or even proposing new QoS aware ones. However key requirements for inter-domain routing like scalability, intra-domain independence, and policy awareness are missing in most of existing multicast routing protocols. In this paper, it is proposed an inter-domain multicast routing strategy that builds inter-domain unidirectional multicast distribution trees, taking into account multicast specific routing policies, and supporting QoS requirements.

O Multicast é um modelo de comunicação onde a transmissão de dados é feita simultaneamente para múltiplos receptores e por isso particularmente útil para suportar aplicações que envolvem vários intervenientes, de que são exemplo, a vídeo-conferência, o ensino-à-distância, o trabalho cooperativo, etc. Apesar de ser um tópico com crescente interesse, onde todos os dias surgem novas propostas, ainda não existem soluções passíveis de serem genericamente adoptadas pelos Fornecedores de Serviço IP. Umas das principais razões apontadas para este facto é a complexidade da maioria das soluções, que atrasa de forma drástica o desenvolvimento alargado dos protocolos de encaminhamento multicast. Neste contexto, o recurso à simulação é particularmente útil. A simulação pode ajudar não só a configurar correctamente alguns dos parâmetros em causa, mas também conduzir à proposta de novas soluções. Neste artigo é apresentada uma implementação do PIM-SM que integramos no Network Simulator (NS). O NS é uma ferramenta muito usada na simulação e teste de protocolos de comunicações. Por sua vez o PIM-SM é um dos protocolos de encaminhamento multicast que está mais desenvolvido na Internet e, como tal, faz sentido estudar as suas limitações com vista à proposta de soluções que as ultrapassem, ou pelo menos, minimizem.

The goal of Differentiated Service Architecture is to provide the benefits of different CoS levels while avoiding the limitations of the IntServ model. This is accomplished by aggregating traffic into specific classes, thus changing the scope from QoS (per flow) to CoS (per class) guarantees. In presence of DiffServ networks, per flow path computation is not adequate. Instead, per class path calculation must be made and so multiple paths (unicast routing) or trees (multicast routing), must be computed in order to satisfy the different QoS requirements of different traffic classes. This paper presents a new multicast routing strategy enabling per class multicast tree computation. The proposed heuristics enable directed trees establishment, instead of reverse path ones, due to the importance of link asymmetry within an environment which is, essentially, unidirectional.